You have the give your drivers a reason to use your app instead of uber.
You have to give riders a reason to use your app instead of uber.
The two reasons have to be good enough to build a large enough network for the rideshare system to work.
I'd say there is a network effect for ride-sharing in aggregate, neither Lyft nor Ueber has been good at locking drivers and riders in with loyalty programs to try to keep people solely on their platform.
The truth is that all the viable competitors are likely already out there and it's going to be very very difficult to replicate what these incumbents have without burning ridiculous amounts of capital.
If the Softbank deal closes, we'll likely see a wave of consolidation with Softbank orchestrating acquisitions by Uber. It's very common for relatively young markets to consolidate over time.
No one in their right mind would start a new company to compete with the existing TNCs in 2017.
Ride sharing is essentially a local business with very low barriers to entry. You don't need to replicate what the incumbents have right away. You just need a modest number of drivers (who can also be driving for the incumbents) and a modest number of customers.
That seems sufficient to me. It's hard to name any company started in the last few years that is a major market player when then have a bunch of competitors. New businesses take time to grow.
The discussion here is about whether Uber has a moat. The point being made is that it's pretty easy to start a rideshare company. Which it is. That one hasn't become major yet isn't proof that one can't become so. As long as there's room for new companies to start and be sustainable at modest scale, then Uber can't extract monopoly rents.
Did the journalist who wrote that story cite any figures to support that assertion?